Executive Summary: Jeremiah (English → French)

Why it matters: Jeremiah’s core passage — the New Covenant (31:31-34) — sits on top of the single most acute lexical collision in this entire language package: the standard French phrase for Jeremiah’s covenant promise, “nouvelle alliance,” is also the ordinary French name for the whole New Testament canon. Combined with two other Critical-risk messianic terms in chapters 23 and 33, and a book-length covenant-lawsuit genre addressed to a French readership split between lapsed/practicing Catholics, a small Protestant minority, and a secularizing unchurched majority, this curriculum carries a denser concentration of Critical and High doctrinal risk than the Romans baseline it extends.

---

Key findings

• Full 52-chapter coverage completed. Every chapter has been analyzed for load-bearing theological vocabulary; chapters carrying no new doctrinal content (5, 12, 14–15, 21, 26, 35, 37–43, 45, 52) are explicitly logged as reviewed, not omitted.
• 66 terms are now recorded in the Jeremiah-specific translation_memory.json: 15 inherited exactly from the Romans baseline (with risk re-evaluated where Jeremiah’s genre intensifies it — e.g., “covenant” rises from Medium to High) and 51 new terms required by Jeremiah’s distinct Hebrew vocabulary (idolatry-as-adultery, the OT three-term sin family, the tetragrammaton register, the righteous Branch, and more).
• A new register decision was required and ratified: the tetragrammaton (YHWH, ~175 occurrences) is fixed as “le SEIGNEUR” — aligning with the baseline’s own “Seigneur” convention — explicitly rejecting the archaic Segond-1910 “l’Éternel” the baseline’s tone rules already forbid.
• A genuine LXX/MT versional divergence at Jeremiah 31:32 (“I was their husband” vs. Hébreux 8:9’s “I disregarded them”) requires a standing translator note rather than harmonization — the first place this curriculum must explicitly protect a textual difference from being “corrected” away.

---

Risks

• Term-level: the Jeremiah-specific term registry records 5 Critical-risk terms (new_covenant, god, yhwh_name, righteous_branch, yhwh_tsidkenu) and 16 High-risk terms (covenant, i_was_their_husband, righteousness, justice_social, everlasting_love, covenant_loyalty, iniquity, idol_idolatry, spiritual_adultery, false_prophet, false_peace, calling, law, return_repent, remnant, plans_of_welfare) — 21 terms total requiring human theologian review.
• Doctrine-level: the doctrine risk registry records 3 Critical doctrines (The New Covenant; Messianic Promise: The Righteous Branch; Deity of the Messianic Branch/YHWH Tsidkenu) and 10 High doctrines, for 13 of 24 doctrines (54%) requiring mandatory human theologian review — nine more require native speaker review, and only two are automated-only.
• France-specific ecumenical exposure: Jeremiah’s idol polemic (ch. 10, throughout) risks being read as implicit anti-Catholic commentary on statuary and Marian devotion unless every occurrence carries an explicit worship-vs-devotional-use gloss.
• France-specific supersessionism exposure: given France’s large Jewish community, “nouvelle alliance” and “maison d’Israël” language must never imply Israel’s replacement or erasure.
• Decontextualization exposure: Jeremiah 29:11 (“des projets de paix”) is flagged as the book’s highest-likelihood devotional misquotation risk, parallel to Romans 8:28 in the baseline.

---

Opportunities

• Jeremiah’s lament material (the “Weeping Prophet” doctrine) lands unusually well in French literary and liturgical culture (Lamentations’ Holy Week/Tenebrae use, the French mystical-confessional tradition) — the lowest-friction doctrine in the curriculum, requiring only native speaker (not theologian) review.
• France’s robust secular vocabulary around sectes/gourous gives “faux prophète” strong, ready-made contemporary resonance for the False Prophets doctrine, provided the text’s own chapter 23 discernment criterion is taught alongside it rather than replaced by it.
• The covenant-ring resonance of French “alliance” (also the word for a wedding ring) is a genuine asset for teaching the relational, not merely legal, force of berit throughout the book.

---

Recommended actions

1. Ratify the “le SEIGNEUR” tetragrammaton register decision before any Phase 2 segment translation begins — it recurs in nearly every chapter and cannot be revised mid-stream without a full re-pass.
2. Route all 13 Critical/High doctrines and 21 Critical/High terms through mandatory human theologian review, prioritizing chapters 23, 29, 31, and 33 for reviewer scheduling given their density of Critical-risk material.
3. Mandate the standing glosses identified in this package (idol/devotional-image distinction; spiritual-adultery metaphor flag; false-peace vs. peace-with-God contrast; New Covenant vs. New Testament canon-title distinction) in every lesson unit where the relevant term occurs — these are not one-time footnotes but recurring, per-occurrence requirements.
4. Carry the LXX/MT versional note at Jeremiah 31:32 into every cross-reference touching Hébreux 8:9, and flag any chapters 46–51 citation of LXX-ordered secondary sources.
5. Load assets/translation_memory.json (Jeremiah) alongside the Romans baseline TM, not in place of it, for every Phase 2 session, per the updated analysis/12_ai_translation_requirements.md.